[Trainer] Move logic for checkpoint loading into separate methods for easy overriding#17043
Merged
sgugger merged 1 commit intohuggingface:mainfrom May 2, 2022
Merged
Conversation
|
The documentation is not available anymore as the PR was closed or merged. |
sgugger
approved these changes
May 2, 2022
Collaborator
sgugger
left a comment
There was a problem hiding this comment.
Thanks for the PR! This will also make the code of the train method easier to read.
Note that you might have to keep up to date with some of the changes we do in those methods (I actually have a bug to fix to resume from sharded chekpoints that will make some changes in your two new private methods) when you subclass in your Trainer.
stevhliu
pushed a commit
to stevhliu/transformers
that referenced
this pull request
May 3, 2022
elusenji
pushed a commit
to elusenji/transformers
that referenced
this pull request
Jun 12, 2022
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR does a small refactoring in the
Trainerclass, specifically it moves the logic for the following two steps out of the training loop into separate helper methods:_load_from_checkpoint()method._load_best_model()method.The PR does not change any existing logic in any way.
Motivation
In our library, we implement a custom Trainer class that subclasses your great built-in Trainer class. However, as we don't save full model checkpoints during training, the mentioned steps for checkpoint loading are not applicable to our use case. Moving this logic to separate methods would be super helpful to us (and potentially others), since we could easily override these helper methods without modifying the training loop itself.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@sgugger
(cc @hSterz)